The AI Journal

In the winter of 1816 into 1817, bread in Basel did not become notably dearer. Sixty miles east, in the textile cantons of St Gallen and Appenzell, people were eating grass pulled for livestock and burying the dead of a famine severe enough that contemporary accounts describe households surviving on bran, nettles and bark. Both places sat inside the same small confederation, under the same stratospheric veil, through the same two ruined harvests. The climate shock that produced them was, for a country two hundred kilometres across, about as close to uniform as a historical climate event gets. The outcome was not.

Mount Tambora, on the Indonesian island of Sumbawa, erupted in April 1815 in the largest explosive eruption directly observed in recorded history. Clive Oppenheimer’s 2003 survey of its climatic, environmental and human consequences traces the tens of megatonnes of sulphate it injected into the stratosphere to a measurable drop in hemispheric temperature that persisted for more than two years, producing the failed harvests of 1816 across Europe and North America that contemporaries called the Year Without a Summer. The aerosol veil that caused this did not respect national borders, and across a country the size of Switzerland it did not meaningfully respect cantonal ones either: the cooling that reached Basel reached Appenzell within the same season.

John D. Post’s 1977 study of the resulting crisis, The Last Great Subsistence Crisis in the Western World, supplies the obvious first explanation for why some places suffered more than others: a common shock does not strike every economy with equal force. Post’s demographic reconstruction finds mortality rising only a few percentage points in France and Prussia through 1817, against an increase of more than a fifth above the normal rate in Switzerland and Tuscany. That gap is usually read as a gap in exposure, the places whose harvests failed worst or whose relief systems were weakest suffering worst. Applied within a single small country, the exposure story runs out of room to work. Switzerland’s cantons in 1816 shared a currency, a climate and two consecutive ruined harvests. If one uniform shock produced wildly uneven outcomes even there, something other than the weather was doing the sorting.

Daniel Krämer’s 2015 study of the crisis, developed from his University of Bern dissertation, supplies that mechanism, and it is not meteorological. Krämer has argued, including in his own presentations of the research, that the cantons’ differing economic orientation, together with their refusal to cooperate with one another, did more to determine the crisis’s severity than the harvest itself. The eastern cantons that suffered worst, St Gallen and Appenzell above all, had spent the preceding century turning themselves into Europe’s leading linen and cotton export region: land that had once grown grain now grew flax, or stood as pasture feeding a rural population of weavers and embroiderers who held no farmland and no grain store of their own. These households did not simply fail to grow enough food in 1816. They had already, deliberately, and for sound economic reasons in an ordinary year, stopped growing most of it.

That arrangement works only for as long as two things hold: that textiles can still be sold for enough to buy grain, and that grain can still be bought. In 1816 and 1817 neither held. Export markets for Swiss cotton goods, already strained by the trade disruptions of the preceding decade, left the weaving cantons with less purchasing power exactly as grain prices rose sharply across the country, tripling in the west and rising sixfold in the east, by Krämer’s own account of the price record. And Switzerland in 1816 was not yet a customs union: each canton still governed its own trade, including the right to stop its own grain leaving its own territory. The lowland cantons that still had a surplus responded to the shortage by embargoing it, keeping grain inside their own jurisdiction rather than letting it cross into the mountains and the east, where it was needed most. The result was a famine shaped less by where the harvest failed than by where a neighbouring government decided not to sell.

Patrick Webb’s 2002 account of Europe’s emergency relief in 1817 situates this within a wider pattern: across the continent, the crisis pushed governments that had never systematically fed their own populations into ad hoc intervention, grain purchases, soup kitchens and price controls, measures contemporaries treated as emergency expedients but that left behind a precedent for state responsibility in subsistence crises that outlasted the famine itself. Switzerland’s cantons absorbed that lesson unevenly and late, precisely because the crisis there was not simply a supply shock waiting on relief but a trade dispute between jurisdictions with no standing mechanism for resolving one: a confederation rather than a state, with no central authority empowered to order a lowland canton to release its grain to a starving neighbour three valleys away.

The strongest objection to reading this as a story about trade rather than climate is that Switzerland’s internal geography is not actually uniform, whatever its area on a map. St Gallen and Appenzell sit higher and further into the foothills than Basel; a cooling event that shortened the growing season by two or three weeks could plausibly have cost the mountain cantons a harvest that would have survived at lower altitude regardless of what anyone exported. On this reading, altitude, not textiles, explains the gap, and the trade story is decoration on a climate story adequate on its own. The objection is reasonable, and it is also incomplete, because it cannot account for the one feature of the crisis with no climatic analogue at all: the embargo. An early frost does not observe a cantonal border, but a magistrate’s order not to sell grain across one does. Basel’s bread prices stayed low not because Basel was warmer than Appenzell in any way that mattered to its 1816 harvest, but because Basel sat on trade routes that could still pull grain in from outside Switzerland altogether, while the mountain cantons depended on neighbours within it who had just decided to keep their own stores for themselves. Altitude can explain why the eastern harvest failed worse. It cannot explain why the resulting deficit was met with an export ban rather than a market.

What a single volcano’s stratospheric veil could not do, discriminate between one Swiss canton and the next, a grain embargo could do easily, and did. The textile cantons’ hunger in 1816 was prepared a century earlier, the moment their fields went over to flax; it was delivered in 1817, the moment a lowland council decided that a bad harvest two valleys over was not its problem. Tambora supplied the shock. Switzerland supplied the unevenness, out of an economy that had already wagered its food supply on exports, and a confederation whose members, when the wager went bad, owed each other nothing.